Re-balancing Asset Allocation during ER

[rwc356]

Should re-balancing your Asset Allocation be conducted in the same fashion during retirement as compared to during the period of saving for retirement?

Should the frequency be the same? Reading the "Buckets" articles - it seems like annual re-balancing of the allocation is the norm.

Is there a better or worse time during the year to re-balance?

Appreciate any thoughts or suggestions.

Bob

 

[LOL!]

I see no reason why rebalancing during retirement is any different than rebalancing while not in retirement.

I do not think annual rebalancing is the norm. The norm is to set ranges and boundaries and rebalance when you go out of your range. Also the norm is to use any distributions to help with rebalancing.

http://www.early-retirement.org/forums/f28/asset-allocation-tutorial-31324-4.html#post586209 (see posts #111, #115, ....)

There is also a recent study called about so-called 'opportunistic rebalancing' where you look often to see if your portfolio has any asset class out of its range, and only rebalance when something is out of range.

 

[ziggy29]

I would think it's no different. The allocations might be different, AND a 65-year-old might not rebalance as much as a 40-year-old because the portfolio is likely less volatile, but other than that, yes, it would be done the same way.

There are different ways to rebalance. One is on a regular periodic basis (I believe Bernstein has recommended a 12- to 18-month period between rebalancing).

There's also a form of rebalancing where you automatically rebalance when certain classes are way out of whack relative to its target allocation. Even this can be modified so that you never rebalance an asset class more than (say) every 6-12 months, otherwise you are losing the short-term benefit of "riding" hot asset classes, which tend to stay hot on the short term. (Remember the mantra: In the short term, "hot" classes tend to stay hot, and in the long term they tend to underperform as they revert to the mean.)

 

[Rich_by_the_Bay]

One difference is that post-FIRE you would typically balance by sweeping money from one fund to another, instead of selectively buying into down funds when you have an income and are investing.

But my readings suggest (I'll see if I can find the source) that there is some advantage to avoiding frequent rebalancing; every year or two seems to be the consensus. I plan to wait until something drifts 15% to 20% out of whack (relative to its own starting point). I can see in the recent market that it would be frantic trying to keep everything balanced too often.

 

[LOL!]

.... I can see in the recent market that it would be frantic trying to keep everything balanced too often.

Actually, I found it quite surprising that it's gonna take practically a meltdown to get things to move to where one needs to rebalance.

In the example in the other thread, a 10% drop in stocks would not trigger a rebalance if one used a trigger of 5% out-of whack. Say you wanted 60% stocks, so your range was 55% to 65%. In this example:

Let’s do another example also starting from 60:40. Suppose stocks drop 10% and bonds go up 5% (think what happened in the last few months). You now have (60 * 0.9 + 40 * 1.05) = 96 or an overall 4% loss. Your equities are now 56% of your portfolio and your bonds are 44%. If you sell 4% of your total portfolio value out of bonds and use that 4% to purchase equities, you will be back to a 60:40 asset allocation.

you are still in range.

 

[mathjak107]

i rebalance bewtween my cash buckets , bonds bucket and equity bucket by years of spending money in bucket 1 (cash)

 

[ziggy29]

Actually, I found it quite surprising that it's gonna take practically a meltdown to get things to move to where one needs to rebalance.

In the example in the other thread, a 10% drop in stocks would not trigger a rebalance if one used a trigger of 5% out-of whack. Say you wanted 60% stocks, so your range was 55% to 65%. In this example:
you are still in range.

The way I see it, a 5% "drift" on 60% takes it to 57-63%. We're looking at 5% variance from a 60% target, and 5% of 60% is 3%. That's how I've understood this method.

(I've toyed with this strategy before, and when I've done it I put it into a spreadsheet with the minimum "drift" required to trigger rebalance as 12% or $1000 per asset class, whichever is greater. (I'm using ETFs so the trading fees on less than $1000 aren't worth the rebalancing.)

And in reality, no single class takes 60% of most allocations unless you only have one all-equities class. In my allocation, no equity class has more than 12%. And not all of these move in unison, so even when some might trigger a rebalance, others might not.

 

[Rich_by_the_Bay]

Actually, I found it quite surprising that it's gonna take practically a meltdown to get things to move to where one needs to rebalance.

In the example in the other thread, a 10% drop in stocks would not trigger a rebalance if one used a trigger of 5% out-of whack. Say you wanted 60% stocks, so your range was 55% to 65%. In this example:
you are still in range.

The frenzy is more likely to come from different sectors within your equity holdings, rather than stocks to bonds. REITs might lag by 10% one month and be up 12% the next, as an example.

 

[LOL!]

The way I see it, a 5% "drift" on 60% takes it to 57-63%. We're looking at 5% variance from a 60% target, and 5% of 60% is 3%. That's how I've understood this method.

It's OK to see it that way, but see it explained by Larry Swedroe who wrote the following on the diehards forum (Bogleheads :: View topic - Larry: I apologize for asking you this, but...) :

Third, you should add a rebalancing table with hold ranges around the targest

I recommend a 5/25 rule that is explained in The Only Guide book.

You should check for rebalancing at three levels
Stocks to bonds, domestic to int'l and individual asset classes

you should rebalance if you are > 5% from target and don't have to pay ST gains or 5% of the absolute target level

So at 60% equity you would have range of 55 and 65 but for say int'l value at say 10% the ranges would be 7.5% and 12.5%

Hope that helps

I think +-3% can lead to too much rebalancing since equities move by more than 3% many weeks and/or months in a year.

 

[ziggy29]

I think +-3% can lead to too much rebalancing since equities move by more than 3% many weeks and/or months in a year.

The problem with your analysis here is that if an asset class moves 3%, it probably changes its own allocation by nowhere near 3%.

If you have $100,000 in assets in a portfolio and you have a 20% allocation for a particular equity class, you have $20,000 in that class and $80,000 in the rest.

Assuming the other 80% stays flat, a 3% gain in the asset in question brings it to a value of $20,600. That brings its allocation to 20.48% (20,600/100,600) -- only 0.48% higher overall, nowhere near 3%.

 

[LOL!]

You got me!

 

[Nords]

I think +-3% can lead to too much rebalancing since equities move by more than 3% many weeks and/or months in a year.

Hey, the concept is right even if the math isn't.

I don't know if this is the study that Rich is looking for, but Walkinwood linked this for me last month:
FPA Journal - Opportunistic Rebalancing: A New Paradigm for Wealth Managers

A 20% swing seems to be the best combination of long-term cap gains (favorable tax treatment) with minimal trading costs. So as we finish rebalancing, each of our four equity asset classes is going to be about 23%-- which we'll rebalance whenever they wander below 18% or above 28%.

We've let Berkshire Hathaway get up to as high as 37% of our ER portfolio, so 23%+/-20% is progress.

 

[ziggy29]

We've let Berkshire Hathaway get up to as high as 37% of our ER portfolio, so 23%+/-20% is progress.

I suspect a lot of people who use Berkshire as part of their allocation have to do this because they have to sell in $4500+ chunks.

 

[Nords]

I suspect a lot of people who use Berkshire as part of their allocation have to do this because they have to sell in $4500+ chunks.

Well, selling the "A" shares would be a bit of a pain-- precisely the reason that Buffett has never split them-- but I see market makers dealing with bid/ask quotes for one or two "B" shares all the time. Not much different from selling 200 shares of a $25/share stock, but since automation & decimalization I doubt anyone worries about 100-share lots anymore.

And charities/donor-advised funds are quite happy to accept a share or two... It's interesting to note that "B" share daily volume has been rising over the last couple years, just about the time that Buffett started giving it away. It's gone from 13,000/day in early 2006 to over 20,000/day today.

Hopefully people concerned about $4500 chunks would invest in a mutual fund that holds a lot of Berkshire. Similar funds over a decade ago were what led Berkshire to set up the "B" shares in the first place.

 

[TromboneAl]

Once upon a time there was a dirty market timer named Tim. Here is Tim's strategy:

"I start out with a 60/40 asset allocation. If my stocks get to 70%, that shows me that the market run is complete, and so I rebalance back to 60/40. If the allocation gets down to 50/50, I see that as a buy signal since the end of the downturn is near."

 

[DblDoc]

I follow the 5/25 guidelines posted above but I'm accumulating so I buy into the laggards. As RichInTampa pointed out you would be shifting from asset classes in retirement or redirecting dividends/interest. I have seen recommendations for rebalancing vary from the opportunistic version to about every 1 to 3 years. What you choose will likely depend on how much time and energy you want to devote to the process and how you have your withdrawal strategy set up. Remember the biggest determination of your return long term is your AA, everything else is a minor component.

DD